Search

Your search keyword '"Li, Peibo"' showing total 436 results
Start Over Author "Li, Peibo"
436 results on '"Li, Peibo"'

1. Large Language Models for Next Point-of-Interest Recommendation

Author

Li, Peibo, de Rijke, Maarten, Xue, Hao, Ao, Shuang, Song, Yang, and Salim, Flora D.

Subjects
Computer Science - Information Retrieval, Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract

The next Point of Interest (POI) recommendation task is to predict users' immediate next POI visit given their historical data. Location-Based Social Network (LBSN) data, which is often used for the next POI recommendation task, comes with challenges. One frequently disregarded challenge is how to effectively use the abundant contextual information present in LBSN data. Previous methods are limited by their numerical nature and fail to address this challenge. In this paper, we propose a framework that uses pretrained Large Language Models (LLMs) to tackle this challenge. Our framework allows us to preserve heterogeneous LBSN data in its original format, hence avoiding the loss of contextual information. Furthermore, our framework is capable of comprehending the inherent meaning of contextual information due to the inclusion of commonsense knowledge. In experiments, we test our framework on three real-world LBSN datasets. Our results show that the proposed framework outperforms the state-of-the-art models in all three datasets. Our analysis demonstrates the effectiveness of the proposed framework in using contextual information as well as alleviating the commonly encountered cold-start and short trajectory problems.

Published
2024
Full Text
View/download PDF

6. Explainability in Graph Neural Networks: An Experimental Survey

Author

Li, Peibo, Yang, Yixing, Pagnucco, Maurice, and Song, Yang

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence, A.1, I.2.0
Abstract

Graph neural networks (GNNs) have been extensively developed for graph representation learning in various application domains. However, similar to all other neural networks models, GNNs suffer from the black-box problem as people cannot understand the mechanism underlying them. To solve this problem, several GNN explainability methods have been proposed to explain the decisions made by GNNs. In this survey, we give an overview of the state-of-the-art GNN explainability methods and how they are evaluated. Furthermore, we propose a new evaluation metric and conduct thorough experiments to compare GNN explainability methods on real world datasets. We also suggest future directions for GNN explainability.

Published
2022

26. Hybrid phase-change lattice Boltzmann simulation of the bubble nucleation and different boiling regimes of conjugate boiling heat transfer

Author

Zhao, Wandong, Liang, Jianhan, Sun, Mingbo, Cai, Xiaodong, and Li, Peibo

Subjects
Physics - Fluid Dynamics, Physics - Computational Physics
Abstract

Pool boiling characteristics in two computational domains with and without considering conjugate heat transfer (CHT) were numerically simulated by an improved hybrid pseudopotential phase-change lattice Boltzmann method (LBM). The effects of constant temperature boundary condition (BC) with nucleate spots and fluctuant temperature on the boiling process were investigated in detail. It was found that for the computational domain without CHT, the treatment of constant temperature BC with nucleate spots is quite easier to produce film boiling than the temperature BC with small fluctuation. However, the results would be the opposite for the case with CHT. The entire boiling curve from the onset of nucleate boiling to fully developed film boiling was presented using the computational domain considering CHT and constant temperature BC with nucleate spots. The simulated critical heat flux showed an excellent agreement with the existing analytical solutions. Hence, the current hybrid phase-change LBM was quantitatively verified. The highly fluctuant heat flux occurred in the CHF and transition boiling as well as the transverse movement of the bubbles had been observed. Furthermore, the thermal responses inside the heater and heat transfer mechanism in different boiling patterns were also comprehensively studied.

Published
2019

43. Review on spray characteristics of liquid–liquid injectors in liquid rocket engines.

Author

Xie, Yu, Zhang, Jiaqi, Sun, Mingbo, Wu, Jiping, Li, Peibo, An, Bin, Liang, Changhai, Wang, Taiyu, Cheng, Peng, Chen, Jikai, Wang, Jiaoru, and Li, Menglei

Abstract

Impinging-jet injectors, liquid–liquid coaxial swirl injectors, and liquid–liquid pintle injectors are representative liquid–liquid injectors in liquid rocket engines (LRE). For these liquid–liquid injectors, the atomization processes all involve the liquid impingement, including jet–jet, sheet–sheet, and jets/sheet–sheet impingement, respectively. After impingement, a liquid sheet forms and fragments. Based on these similarities, reviewing published literature on the spray characteristics of these three liquid–liquid injectors in LRE is necessary and will facilitate the investigation of spray characteristics of liquid–liquid pintle injectors to meet the progress of variable-thrust LRE. This review covers the following aspects of these injectors: basic spray morphology, liquid sheet characteristics and disintegration mechanisms, and atomization characteristics. For impinging-jet injectors, rim instability and impact wave play crucial roles in spray morphology and disintegration. Jet Weber number is of great importance for liquid sheet breakup length and mean droplet diameter. In the case of liquid–liquid coaxial swirl injectors, the overall spray morphology is similar to that of pressure swirl injectors, but it may feature two separate liquid sheets. The recess length strongly influences spray morphology, spray angle, breakup length, and Sauter mean diameter. Liquid–liquid pintle injectors can be simplified to injection element, in which the spray morphology resembles a cloak-like shape. In a complete pintle injector, the spray forms a conical liquid sheet. Momentum ratio proves to be the most significant parameter for predicting spray angle. Although the review indicates substantial progress has been made in understanding spray characteristics of liquid–liquid injectors, there remain several shortcomings that require further research, particularly for pintle injectors, which can be learned from the other two injectors. [ABSTRACT FROM AUTHOR]

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results